A lot of advice online tells you to split test pretty much everything that you can. Leaps of faith aren’t effective anymore, and they can cause you to exhaust your resources without seeing any return on your investment. By collecting the right information based on real-time user feedback, you can make constructive changes that lead to higher conversions, more sales and increased ROI.

In order to make split testing effective for your marketing strategies, there are a few things to know. Let’s explore.

Refrain from Making Assumptions

It’s easy to make assumptions about your audience, particularly if you’ve been doing this for a long time. Could your customers really have changed that much?

You might think that you know what your customers want or what their pain points are. You may believe that you know their location, their age or their interests. Resist the temptation!

Customer profiling is based on assumptions rather than clear, concrete data. Thanks to advances in technology, you can now measure the impact of every digital touchpoint. This way, you can learn about your customer’s preferences individually and as a whole. You know what assuming does after all!

Establish a Baseline

Another way to ensure success with A/B testing is to have a clearly established baseline. What this means is that you are aware of what your current conversion rates are so that you can compare the progress you’re making.

If you don’t have a baseline, it’s going to be almost impossible to know if the changes you’re making are having an impact or not. Don’t rush through this step, and you’ll be much happier with your results.

Test One Thing at a Time

We know that you’re in a hurry, but don’t try to tackle too many things at once. With A/B testing, you can only test one variable at a time. If you change your message and your CTA, how will you know which one led to more conversions?

Luckily, split testing can be implemented slowly, so you don’t need to have your answers immediately to increase conversions. You may split test your headline, and then expand to your messaging, your keywords, your CTA, your images and so forth.

It’s also worth noting that it’s OK to make mistakes. You’re not perfect, and no one expects you to be. (Except maybe your parents). The key is that you are open and willing to learn from your mistakes and become a better A/B tester.

Realize that Your Results Could be a Fluke

You can’t run one test and assume that the “winner” will always be the winner. In A/B testing, you must think in terms of statistical confidence, which means that you will reach the same results again and again.

Here are some of the elements to think about.

First is sample size. What size sample did you run the test on? If it was only a few people, you may be off to a good start, but you need a larger group to know the true impact.

Second is percentage. The accuracy of your A/B test will depend on your margin for error. If you have a sample size of 100 people and 99 percent converted when shown a larger CTA button, your margin of error is low. However, if you had 52 percent convert with the larger button and 48 percent convert with the smaller button, your margin of error is much higher.

Third is the population size. If your entire audience is 100,000 and your sample size is 100, your results will have a lower confidence level. Practice makes perfect, so you can continue split testing to have a higher degree of confidence.

Create Your Own Success

It’s normal to turn to the internet for advice from other marketers, but do realize that there is no one-size-fits-all solution. If there were, you wouldn’t need split testing in the first place!

Your audience is entirely unique, which means you can generalize as to what will work for you. So feel free to browse ideas from other marketers and play around with their advice, but don’t believe that because it worked for them, it will work for you. Your audience may respond better to a green CTA button over a red one, for instance.

Collect Feedback

We look to second opinions all of the time. From doctors to friends to significant others, we rarely take things at face value. But when it comes to user testing, we often run one test and think that things will always work that way for the rest of time.

Fortunately, you don’t need to have a UX Department on hand to give you the second opinion that you need. There are a handful of free or affordable services that allow you to perform usability testing in a pinch.

Some of our favorites include Amazon Mechanical Turk and Peek User Testing. They aren’t perfect, and there are limitations, but you can at least gather some feedback that can help with your testing.

Define Your Success Metric

A final tip for split testing for the best results is to clearly define your success metric. A/B testing is about finding what increases conversions – not about open rates, click-through rates, shares, etc. It’s easy to get caught up in these metrics, but the true sign of success is getting users to convert.

As you build out your campaigns, make sure that you are clear about your goal and that everything is optimized to achieve that goal.


Don’t be afraid to run A/B testing on your entire website. It’s a cost-effective way to learn about what’s working for your audience. Just make sure that you don’t lose sight of the real objective, which is delivering the right content to the right audience at the right time. It’s not just about getting clicks. You must be willing to look at your site objectively and implement the appropriate changes based on your data – not assumptions.